Skip to content

[FRT-1629] Dedupe retryable statement errors and always surface a failure reason #3400

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Draft
Natea Eshetu Beshada (nateab) wants to merge 1 commit into
base: main
Choose a base branch
from
Draft

[FRT-1629] Dedupe retryable statement errors and always surface a failure reason #3400

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
35 changes: 32 additions & 3 deletions pkg/flink/internal/store/store.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -128,9 +128,14 @@ func (s *Store) WaitPendingStatement(ctx context.Context, statement types.Proces
// Check for failed or cancelled statements
statementStatus = updatedStatement.Status
if statementStatus != types.COMPLETED && statementStatus != types.RUNNING {
failureMessage := updatedStatement.StatusDetail
if failureMessage == "" {
failureMessage = fmt.Sprintf("the server did not report a failure reason; run `confluent flink statement describe %s` for the full status", statement.StatementName)
}
return nil, &types.StatementError{
Message: fmt.Sprintf("can't fetch results. Statement phase is: %s", statementStatus),
FailureMessage: updatedStatement.StatusDetail,
FailureMessage: failureMessage,
Suggestion: "This statement reached a terminal state without producing results. Check the statement definition and run `confluent flink statement describe` for the full status.",
StatusCode: types.StatusCode(err),
}
}
Expand Down Expand Up @@ -249,7 +254,7 @@ func (s *Store) waitForPendingStatement(ctx context.Context, statementName strin
return nil, &types.StatementError{
Message: fmt.Sprintf("the server can't process this statement right now, exiting after %d retries",
len(capturedErrors)),
FailureMessage: fmt.Sprintf("captured retryable errors: %s", strings.Join(capturedErrors, "; ")),
FailureMessage: formatCapturedErrors(capturedErrors),
}
}

Expand Down Expand Up @@ -277,7 +282,7 @@ func (s *Store) waitForPendingStatement(ctx context.Context, statementName strin

var errorsMsg string
if len(capturedErrors) > 0 {
errorsMsg = fmt.Sprintf("captured retryable errors: %s", strings.Join(capturedErrors, "; "))
errorsMsg = formatCapturedErrors(capturedErrors)
}

return nil, &types.StatementError{
Expand All @@ -287,6 +292,30 @@ func (s *Store) waitForPendingStatement(ctx context.Context, statementName strin
}
}

// formatCapturedErrors collapses repeated retryable status details into a single,
// de-duplicated, insertion-ordered summary so an identical error is not printed once
// per retry.
func formatCapturedErrors(capturedErrors []string) string {
counts := make(map[string]int, len(capturedErrors))
order := make([]string, 0, len(capturedErrors))
for _, e := range capturedErrors {
if _, seen := counts[e]; !seen {
order = append(order, e)
}
counts[e]++
}

parts := make([]string, 0, len(order))
for _, e := range order {
if counts[e] > 1 {
parts = append(parts, fmt.Sprintf("%s (repeated %d times)", e, counts[e]))
} else {
parts = append(parts, e)
}
}
return fmt.Sprintf("captured retryable errors: %s", strings.Join(parts, "; "))
}

func (s *Store) getStatusDetail(statementObj flinkgatewayv1.SqlV1Statement) string {
status := statementObj.GetStatus()
if status.GetDetail() != "" {
Expand Down
7 changes: 6 additions & 1 deletion pkg/flink/internal/store/store_onprem.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -117,9 +117,14 @@ func (s *StoreOnPrem) WaitPendingStatement(ctx context.Context, statement types.
// Check for failed or cancelled statements
statementStatus = updatedStatement.Status
if statementStatus != types.COMPLETED && statementStatus != types.RUNNING {
failureMessage := updatedStatement.StatusDetail
if failureMessage == "" {
failureMessage = fmt.Sprintf("the server did not report a failure reason; run `confluent flink statement describe %s` for the full status", statement.StatementName)
}
return nil, &types.StatementError{
Message: fmt.Sprintf("can't fetch results. Statement phase is: %s", statementStatus),
FailureMessage: updatedStatement.StatusDetail,
FailureMessage: failureMessage,
Suggestion: "This statement reached a terminal state without producing results. Check the statement definition and run `confluent flink statement describe` for the full status.",
StatusCode: types.StatusCode(err),
}
}
Expand Down
74 changes: 72 additions & 2 deletions pkg/flink/internal/store/store_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,6 @@ import (
"fmt"
"net/http"
"reflect"
"strings"
"testing"
"time"

Expand Down Expand Up @@ -244,7 +243,7 @@ func TestWaitForPendingHitsErrorRetryLimit(t *testing.T) {
}
expectedError := &types.StatementError{
Message: "the server can't process this statement right now, exiting after 6 retries",
FailureMessage: fmt.Sprintf("captured retryable errors: %s", strings.Repeat(testStatusDetailMessage+"; ", 5)+testStatusDetailMessage),
FailureMessage: fmt.Sprintf("captured retryable errors: %s (repeated 6 times)", testStatusDetailMessage),
}
client.EXPECT().GetStatement("envId", testStatementName, "orgId").Return(statementObj, nil).AnyTimes()
processedStatement, err := s.waitForPendingStatement(context.Background(), testStatementName, timeout)
Expand Down Expand Up @@ -2077,6 +2076,7 @@ func (s *StoreTestSuite) TestWaitPendingStatementFailsOnNonCompletedOrRunningSta
expectedError := &types.StatementError{
Message: "can't fetch results. Statement phase is: FAILED",
FailureMessage: testStatusDetailMessage,
Suggestion: "This statement reached a terminal state without producing results. Check the statement definition and run `confluent flink statement describe` for the full status.",
}

{ // Cloud store
Expand Down Expand Up @@ -2144,6 +2144,7 @@ func (s *StoreTestSuite) TestWaitPendingStatementFetchesExceptionOnFailedStateme
expectedError := &types.StatementError{
Message: "can't fetch results. Statement phase is: FAILED",
FailureMessage: exception1,
Suggestion: "This statement reached a terminal state without producing results. Check the statement definition and run `confluent flink statement describe` for the full status.",
}

{ // Cloud store
Expand Down Expand Up @@ -2220,6 +2221,75 @@ func (s *StoreTestSuite) TestWaitPendingStatementFetchesExceptionOnFailedStateme
}
}

func (s *StoreTestSuite) TestWaitPendingStatementSetsPlaceholderWhenFailedWithNoReason() {
expectedError := &types.StatementError{
Message: "can't fetch results. Statement phase is: FAILED",
FailureMessage: fmt.Sprintf("the server did not report a failure reason; run `confluent flink statement describe %s` for the full status", testStatementName),
Suggestion: "This statement reached a terminal state without producing results. Check the statement definition and run `confluent flink statement describe` for the full status.",
}

{ // Cloud store
client := mock.NewMockGatewayClientInterface(gomock.NewController(s.T()))
store := Store{
Properties: NewUserPropertiesWithDefaults(map[string]string{"TestProp": "TestVal"}, map[string]string{}),
client: client,
appOptions: &types.ApplicationOptions{
OrganizationId: "orgId",
EnvironmentId: "envId",
},
tokenRefreshFunc: tokenRefreshFunc,
}

statementObj := flinkgatewayv1.SqlV1Statement{
Name: flinkgatewayv1.PtrString(testStatementName),
Status: &flinkgatewayv1.SqlV1StatementStatus{
Phase: "FAILED",
},
}

client.EXPECT().GetStatement("envId", testStatementName, "orgId").Return(statementObj, nil)
client.EXPECT().GetExceptions("envId", testStatementName, "orgId").Return([]flinkgatewayv1.SqlV1StatementException{}, nil)

processedStatement, err := store.WaitPendingStatement(context.Background(), types.ProcessedStatement{
StatementName: testStatementName,
Status: types.PENDING,
})
require.Nil(s.T(), processedStatement)
require.Equal(s.T(), expectedError, err)
}
{ // On-prem store
client := mock.NewMockCmfClientInterface(gomock.NewController(s.T()))
store := StoreOnPrem{
Properties: NewUserPropertiesWithDefaults(map[string]string{"TestProp": "TestVal"}, map[string]string{}),
client: client,
appOptions: &types.ApplicationOptions{
EnvironmentId: "envId",
},
tokenRefreshFunc: tokenRefreshFunc,
}

statementObj := cmfsdk.Statement{
Metadata: cmfsdk.StatementMetadata{
Name: testStatementName,
},
Status: &cmfsdk.StatementStatus{
Phase: "FAILED",
},
}

client.EXPECT().CmfApiContext().Return(context.Background()).Times(2)
client.EXPECT().GetStatement(context.Background(), "envId", testStatementName).Return(statementObj, nil)
client.EXPECT().ListStatementExceptions(context.Background(), "envId", testStatementName).Return(cmfsdk.StatementExceptionList{}, nil)

processedStatement, err := store.WaitPendingStatement(context.Background(), types.ProcessedStatement{
StatementName: testStatementName,
Status: types.PENDING,
})
require.Nil(s.T(), processedStatement)
require.Equal(s.T(), expectedError, err)
}
}

func (s *StoreTestSuite) TestGetStatusDetail() {
exception1 := "Exception 1"
exception2 := "Exception 2"
Expand Down